<html><head><style type='text/css'>p { margin: 0; }</style></head><body><div style='font-family: Times New Roman; font-size: 12pt; color: #000000'>Dear Nenad,<br><br>You're welcome of course. You're right that after head movement compensation using ft_regressconfound, the sensor level data ideally is not used anymore for source modeling (see http://fieldtrip.fcdonders.nl/example/how_to_incorporate_head_movements_in_meg_analysis -> bottom page, for why this is problematic), i.e. as is required for ft_megrealign. <br><br>Optimally, ft_regressconfound is therefore used as a last step just prior to ft_XXXstatistics, whether at the sensor level (after ft_megrealign) or at the source level. To address your objective, i.e. showing that a difference is due to the experimental manipulation and not due to different head positions; using ft_regressconfound on trials of, say, condition A and B combined will remove any trial-by-trial signal variance that is due to different head positions from the mean head position of condition A and B. In other words, if a systematic difference observed between condition A and B is caused by systematically different head position between condition A and B, this difference will be removed by ft_regressconfound. If not, the difference may not be caused by different head position, and you have good indication to exclude head position as a potential confound. <br><br>Yours,<br>Arjen <br><br><br><hr id="zwchr"><blockquote style="border-left:2px solid rgb(16, 16, 255);margin-left:5px;padding-left:5px;"><b>Van: </b>"Nenad Polomac" <polomacnenad@gmail.com><br><b>Aan: </b>fieldtrip@science.ru.nl<br><b>Verzonden: </b>Vrijdag 12 juli 2013 15:43:35<br><b>Onderwerp: </b>Re: [FieldTrip] ft_megrealign how to avoid?<br><br>Dear Arjen,<br><br>Thank you very much for you answer! I forgot to mention that I applied ft_megrealign on averaged single subject data since we are interested in evoked auditory early gamma response. I am familiar with the online and offline methods for removing variance which comes from head movement. These are very useful tolls. I've tested ft_regresconfound before and it worked fine. However, it will not be very helpful in this case. Because if I apply ft_regresconfound as you suggested than I will loose gradiometers information and than ft_megrealign will not work. Furthermore my evoked gamma grand average topographies (TFR of planar and axial gradiometers) look now as expected without any transformation(ft_regresconfound or ft_megrealign). So what I need is some objective evidence that condition difference is due to experiment rather than different head positions. :)<br>
<br>Thank you and all the best!<br><br>Nenad
<br>_______________________________________________<br>fieldtrip mailing list<br>fieldtrip@donders.ru.nl<br>http://mailman.science.ru.nl/mailman/listinfo/fieldtrip</blockquote><br></div></body></html>